feat(init): auto-populate ignore patterns from detected ecosystem#53
Merged
Conversation
codegraph PR reviewDiff vs main · severity ≤
|
The `unresolved_edges` count conflated two very different things: 1. External imports — `fastapi`, `os`, `react`. Static analysis without a venv/site-packages parse can't resolve these and isn't supposed to. These are expected, not errors. 2. Local-unresolved — target's root module IS in the repo but the specific symbol couldn't be matched (rename drift, dynamic import, missed binding). These are the actionable ones. `compute_metrics` now classifies each unresolved edge by checking whether the target's root module appears as a MODULE node's qualname-root in the graph. New fields: `external_edges`, `unresolved_local_edges`. The `unresolved_edges` total is kept as the sum for back-compat. Surfaced + exposed in the markdown report and the MCP `metrics` tool. v0.1.2 backlog item #2.
FastAPI Pydantic request-body models were being flagged dead because the only reference to them came through a handler's parameter type annotation, which the Python parser didn't trace. The parser now walks each function's parameter `type` strings and the return-type string, regex-extracts capitalized identifiers, drops typing-scaffolding (`Optional`, `Annotated`, `Union`, ...) and FastAPI dependency markers (`Body`, `Depends`, ...), and emits one `CALLS` edge with `dst="unresolved::<TypeName>"` and `metadata.via == "annotation"` per surviving name. The existing resolver rewrites these to real `CLASS` ids when the type is in-repo, which closes the loop for dead-code detection. Framework-agnostic: every function with type annotations benefits, not just route handlers. The change reuses `EdgeKind.CALLS` (rather than introducing a new `REFERENCES` kind) because `CALLS` already flows through the resolver and the reference-kind whitelist used by dead-code, and Pydantic-body validation IS effectively a runtime call into the model class. v0.1.2 backlog item #3.
Coding agents (Claude Code, Cursor, Codex) defaulted to grep even with polycodegraph's MCP server registered, because they had no hint that the MCP tools existed. `codegraph init` now prompts (default yes) and writes a `## polycodegraph` section telling the agent which MCP tool maps to each kind of structural query. Mirrors the `.mcp.json` behaviour: creates fresh, appends when missing, leaves alone when the section is already there. Both CLAUDE.md and AGENTS.md are written. v0.1.2 backlog item #4.
The "Extra ignore patterns" free-text prompt was a footgun — a user typed `y` thinking it was yes/no and got `ignore: [- y]` in their YAML. Init now sniffs the repo for `pyproject.toml`, `package.json` (with React Native detection via dependencies/devDependencies), `go.mod`, `Cargo.toml`, `pom.xml`, `build.gradle`, and pre-fills the matching ignore patterns. The user confirms with a single yes/no. Extras prompt remains for project-specific paths. New module: `codegraph/init_presets.py` with `detect_ignore_presets`. Patterns auto + extras get de-duplicated in insertion order. v0.1.2 backlog item #5.
b2f8783 to
26b2729
Compare
codegraph PR reviewDiff vs main · severity ≤
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
v0.1.2 backlog item #5.
Test plan
🤖 Generated with Claude Code